no luck with Qt 5.9.2

[Jan Mulder]

On 14-10-17 12:59, Dirk Hohndel wrote:
> On Thu, Oct 12, 2017 at 08:04:03PM +0200, Jan Mulder wrote:
>>
>> That was a useful tip, thanks Lubomir. Not out of the woods yet but an
>> interesting observation:
>>
>> Kirigami.ApplicationWindow {
>> 	id: rootItem
>> 	title: "Subsurface-mobile"
>> }
>>
>> gives same problem, but
>>
>> ApplicationWindow {
>> 	id: rootItem
>> 	title: "Subsurface-mobile"
>> }
>>
>> passes the hard exit. So, there is some relation to Kirigami.  However, all
>> this just does not happen on Qt 5.9.1 but only on 5.9.2.
>>
> 
> Hi Jan,
> 
> Any progress on that? I'm pretty sure this is one of my iOS issues right
> now. Obviously I could download Qt 5.9.1 again (different build system, so
> I don't have the 'old' 5.9.1 copy on this machine...) but that's only a
> short term workaround (and when I tried with 5.9.1 the BLE code would
> crash on my iPad, so I hoped that maybe some of the bugfixes in 5.9.2
> might help)


Unfortunately, very very limited progress. I found out that 2nd code 
fragment above does not tell the whole story. Obviously, it needs an 
"import QtQuick.Controls 2.2" the get the ApplicationWindow defined. 
However, adding "import org.kde.kirigami 2.0 as Kirigami" triggers the 
exit on the "can't create window object". So, its only one input that 
triggers the error of an almost empty main.qml

The big question is now: is it Kirigami or Qt/QML? Of even a combination 
of both?

I even tried to compile mobile-on-desktop against the beta of Qt 5.10.0. 
No luck there, and the exact same behavior.

I am heavily tempted to call this an Qt/QML bug. This is not only 
related to this 5.9.2 problem. Some months ago, when I was trying to get 
the BLE for OSTC to work, I had an Qt/QML android only source tree. I 
could (Android) compile the app perfectly against this (and leading to a 
properly running app), but after I synced my repo to the latest Qt 
master and re-compiled Qt from source again, the Subsurface-mobile app 
was crashing at the very start (but not exactly at the "can't create 
window object" exit), so, at that time, I abandoned that source tree, 
assuming I did something wrong, but in hindsight, I might have pulled in 
the Qt bug. All very speculative at this point, but I do not like the 
idea of bisecting some weird Qt/QML issue between 5.9.1 and 5.9.2.

--jan